Poor maternal nutritional status remains an important determinant of long-term maternal health as well as of fetal growth and subsequent infant health and survival, especially in South Asia. We will conduct 3 experiments among a single group of 5,000 undernourished women who live in Matlab Thana, the well-established field site of International Centre for Diarrhoeal Disease Research, Bangladesh. An on-going surveillance program identifies pregnant women within 6-8 wk of conception. An on-going government program provides a food supplement to pregnant and lactating women that contains 600 kcal/d (6 d/wk) and a daily pill that contains 60 mg iron (Fe) and 400 mcg folic acid. Intervention 1: We will randomly assign women to receive advice to begin the food supplementation program (a) immediately after diagnosis of pregnancy (early care) or (b) at the time of their choosing (usual care). We postulate that those in the early care group will have higher-birthweight infants than those in the usual care group. Intervention 2: Within each of these groups, we will randomly assign women to receive a pill that contains (a) 30 mg Fe and 400 mcg folic acid or (b) 60 mg Fe and 400 mcg folic acid (usual care) or (c) 30 mg Fe, 400 jig folic acid and additional micronutrients. We postulate that the lower-iron treatment will have the same effect on maternal hemoglobin but fewer side-effects and, thus, greater compliance than usual care. We postulate that the micronutrient supplement will increase maternal hemoglobin compared to usual care. Intervention 3: We will randomly assign all of the subjects to receive either (a) counseling for exclusive breastfeeding (EBF) or (b) a different health education message of equivalent intensity. We postulate that those who receive counseling for EBF have a longer duration of EBF than those who did not receive this counseling. We postulate that this treatment will increase infant growth and reduce infant morbidity. Each of these trials is designed to address an important scientific issue and also uses an intervention that could be readily incorporated into public health programs.